search

aristoteleo / dynast-release

Inclusive and efficient quantification of labeling and splicing RNAs for time-resolved metabolic labeling based scRNA-seq experiments
https://dynast-release.readthedocs.io/en/latest/
MIT License
14 stars 4 forks source link

Consensus.bam lost gene-tag after completion #14

Closed mhagemann86 closed 2 years ago

mhagemann86 commented 2 years ago

Hi,

I am trying out your very nice dynast pipeline (version 0.2.0) I ran into an issue with counting after creating a consensus bam file. Somehow the consensus.bam has lost the gene-tag or gene assignment after completion. Mind you the flag for gene-tag differs from your default value "GX", but I specify --gene-tag GE when running the consensus command. I checked the in and output bam files for the presence of my gene-tag (see below), plus here is the command I ran

dynast consensus -t 24 -g gencode.GRCm39.vM29.primary_assembly.annotation.noChr.rRNA.gtf --umi-tag UB --barcode-tag BC --gene-tag GE --strand forward --quality 27 -o dynast_consensus UMIs.bam --barcodes expected_barcodes.txt
samtools view -@ 48 UMIs.bam | grep -c 'GE:Z:ENSMUS'
1997940041

samtools view -@ 48 consensus.bam | grep -c 'GE:Z:ENSMUS'
0

Is this normal and one should disregard the --gene-tag for counting after consensus building, or do you have any good input as to why this might have happened?

Thanks in advance.

Best, Michael

Lioscro commented 2 years ago

Hi, @mhagemann86,

The BAM output from the consensus command should have the gene ID placed in the GX tag and gene name in the GN tag, regardless of what tags were used as input. Would you please check if these tags are present?

mhagemann86 commented 2 years ago

Hi,

Thanks for quick reply. Haha, indeed it seems very much like so (see below)

samtools view -@ 48 consensus.bam | grep -c 'GX:Z:ENSMUS'
101906759

Great, that solved it. Might be good to add a small snippet about it in the documentation, that tags will be changed regardless of input tag.

/Michael